Build AI agents and knowledge bases using visual workflow orchestration, data processing, and retrieval augmented generation.

FastGPT is an open-source application that provides a platform for building AI agents and knowledge bases. It combines data processing and model invocation with a visual workflow editor to create automated support and information retrieval systems. The software is deployed via Docker for self-hosting or can be managed through Sealos Cloud for a more streamlined installation process.
Users interact with the platform through a web interface to orchestrate agent skills and manage documents. It allows for the creation of complex application scenarios by linking different nodes in a visual flow, making it suitable for organizations that need to automate customer service, internal knowledge management, or specialized AI assistants. The platform handles the entire pipeline from raw data ingestion to the final model response.
The system is built using Next.js and supports various large language models such as OpenAI, Claude, DeepSeek, and Qwen. It focuses heavily on Retrieval-Augmented Generation (RAG) by allowing users to import data from URLs or files, which is then processed into searchable chunks. The architecture includes a dedicated debugging suite that allows operators to perform single-point search testing and modify or delete citations during a live conversation to improve accuracy. It also includes operational tools for labeling data and reviewing conversation records to refine the agent's performance over time.
FastGPT is a flexible framework for developers and businesses to deploy private AI agents with integrated data management and visual logic control.
A self-hosted control center for running autonomous coding agents to plan and ship changes across codebases.
A collaborative platform to build, schedule, and operate AI agents that handle long running automated tasks.
Join our newsletter to get shiny new open source software delivered to your inbox. Unsubscribe anytime.